Statement of problem
Data comparing the fatigue performance of adhesively luted glass or polycrystalline ceramic systems for computer-aided design and computer-aided manufacturing (CAD-CAM) are scarce.
The purpose of this in vitro study was to evaluate and compare the fatigue performance of monolithic crowns manufactured from glass or polycrystalline CAD-CAM ceramic systems adhesively luted to a dentin analog.
Materials and Methods
Fifty-four pairs of standardized preparations of dentin analog (NEMA Grade G10) and simplified ceramic crowns of 1.5-mm thickness were obtained with 3 ceramic materials: lithium disilicate (LD) glass-ceramic (IPS e.max CAD); zirconia-reinforced lithium silicate (ZLS) glass-ceramic (Vita Suprinity); and translucent yttrium fully stabilized polycrystalline zirconia (Trans YZ) (Prettau Anterior). The simplified crowns (n=15) were adhesively cemented onto the preparations and subjected to step-stress fatigue test (initial load of 400 N, 20 Hz, 10 000 cycles, followed by 100-N increment steps until failure). Collected data (fatigue failure load [FFL] and cycles for failure [CFF]) were submitted to survival analysis with the Kaplan-Meier and Mantel-Cox post hoc tests (α=.05) and to Weibull analysis (Weibull modulus and its respective 95% confidence interval). Failed crowns were submitted to fractography analysis. The surface characteristics of the internal surface (roughness, fractal dimension) of additional crowns were accessed, and the occlusal cement thickness obtained in each luted system was measured.
Trans YZ crowns presented the highest values of FFL, CFF, and survival rates, followed by ZLS and LD (mean FFL: 1740 N>1187 N>987 N; mean CFF: 149 000>92 613>73 667). Weibull modulus and cement thickness were similar for all tested materials. LD presented the roughest internal surface, followed by ZLS (mean Ra: 226 nm>169 nm>93 nm). The LD and ZLS internal surfaces also showed higher fractal dimension, pointing to a more complex surface topography (mean fractal dimension: 2.242=2.238>2.147).
CAD-CAM monolithic crowns of Trans YZ show the best fatigue performance. In addition, ZLS crowns also showed better performance than LD crowns.
Posterior monolithic crowns made of translucent zirconia are more resistant to mechanical fatigue than glass-ceramics. Moreover, zirconia-reinforced lithium silicate ceramic crowns are more fatigue resistant than lithium disilicate crowns.
Monolithic ceramic restorations produced with computer-aided design and computer-aided manufacturing (CAD-CAM) systems have increased in popularity. Their main advantages are more conservative preparations with a minimally invasive concept, elimination of the risk of veneer chipping, extensively reported in clinical trials, , and a completely digital workflow, providing a precise, efficient, accurate, and less time-consuming manufacturing process.
Three ceramic materials have been advocated for such use in the posterior region: lithium disilicate (LD), zirconia-reinforced lithium silicate (ZLS), and translucent yttrium fully stabilized polycrystalline zirconia (Trans YZ). LD has been reported to have an approximately 98% survival rate in up to 10 years of follow-up. It comprises glass reinforced by LD crystals and, after etching with hydrofluoric acid and the application of a silane coupling agent, has adequate adhesion to tooth substrate by using resin luting agents. , ZLS is a recently developed alternative to LD and is classified as a glass-ceramic reinforced by lithium silicate and zirconia crystals within its glassy matrix. Studies have reported a better than or similar performance to LD. , Trans YZ has improved esthetic properties over traditional zirconia, achieved by increasing the stabilizer oxide content and grain size , to provide a complete polycrystalline microstructure. However, such esthetic enhancement reduces mechanical properties, and the flexural strength of translucent zirconia is similar to that of LD and ZLS glass-ceramics. , ,
Studies comparing the fatigue performance of these different systems with similar clinical indication are scarce, and whether these different materials have distinct clinical indications is unclear. For example, Trans YZ has been advocated as being less abrasive to opposing tooth structure and also for situations where it is necessary to mask discolored substrate. LD has been reported to be more esthetically predictable, with excellent mimicking characteristics similar to those of natural tooth dentition ; however, data regarding the performance of ZLS are scarce.
The mechanical performance of a ceramic restoration can be improved with optimal adhesive resin cementation if the irregularities are completely filled by the resin luting agent. Surface treatments should be specifically selected for each ceramic material by considering its microstructure to enhance micromechanical and chemical interactions, even more important with aging. Glass-ceramics have been reported to provide improved fatigue resistance than translucent zirconia, which is a difficult material to bond to. ,
The purpose of this in vitro study was to evaluate and compare the fatigue performance of monolithic crowns manufactured from glass (LD or ZLS) or polycrystalline (Trans YZ) CAD-CAM systems adhesively luted to a dentin analog substrate. In addition, the surface characteristics of the internal surface (roughness, fractal dimension) of crowns were accessed, and the occlusal cement thickness obtained in each luted system was measured. The null hypothesis was that no difference would be found among the evaluated materials.
Material and Methods
|Material||Commercial Name/Manufacturer||Composition||Batch Number|
|Dentin analog||NEMA Grade G10; Accurate Plastics Inc||Continuous ﬁlament woven ﬁberglass bonded with epoxy resin||-|
|Lithium disilicate||IPS e.max CAD; Ivoclar Vivadent AG||SiO 2 , Li 2 O, K 2 O, P 2 O 5 , ZrO 2 , ZnO, and other coloring oxides||W12668|
|Lithium silicate-reinforced zirconia||VITA Suprinity; Vita Zahnfabrik||SiO 2 ; Li 2 O; K 2 O; P 2 O 5 ; ZrO 2; Al 2 O 3; CeO 2 ; pigments||48942|
|Translucent yttrium fully stabilized polycrystalline zirconia||Prettau Anterior; Zirkonzahn SRI|| ZrO 2 ; Y 2 O 3 ; Al 2 O 3 ; SiO 2 ; FE 2 O 3 ; Na 2 O;
|Hydrofluoric acid||IPS Ceramic Etching Gel; Ivoclar Vivadent AG||5% concentration hydrofluoric acid (main component)||W14921|
|Aluminum oxide||Polidental Indústria e Comércio||45-μm Al 2 O 3 particles||44493|
|Primers||Multilink Primer A; Ivoclar Vivadent AG||Aqueous solution of initiators||W89775|
|Multilink Primer B; Ivoclar Vivadent AG||HEMA, phosphate acidic monomers, polyacrylic acid, stabilizers.||W16102|
|Silane||Monobond Plus; Ivoclar Vivadent AG||4% adhesive monomers, 96% ethanol||W10892|
|Resin-based luting agent||Multilink Automix; Ivoclar Vivadent AG||Bis-EMA ethoxylate, UDMA, Bis-GMA, HEMA, barium-glass, ytterbium trifluoride, mixed spheroidal oxides, stabilizer, catalyst and pigments.||W30149|
|Groups||Ceramic Material||Surface Conditioning||Luting Agent||Analysis Performed (Sample Size)|
|LD||Lithium disilicate glass-ceramic (IPS e.max CAD)||5% Hydrofluoric acid applied for 20s, followed by silane||Dual polymerizing resin cement|| Fatigue (n=15)
|ZLS||Zirconia reinforced-lithium silicate glass-ceramic (VITA Suprinity)|
|Trans YZ||Translucent yttrium fully stabilized polycrystalline zirconia (Prettau Anterior)||Airborne-particle abrasion with aluminum oxide particles (45 μm) for 15 s, followed by silane|
Fifty-four standardized preparations for monolithic crowns were produced (A25; Ergomat) from a dentin analog (NEMA Grade G10; Accurate Plastics), as described in the study by Schestatsky et al. A simplified crown geometry with a flat occlusal surface was used, with a final thickness of 1.5 mm for all regions (occlusal surface, axial walls, and chamfer region) and axial walls that converged by 8 degrees. One preparation was scanned (S600 ARTI; Zirkonzahn), and simplified restorations were milled (M1; Zirkonzahn) from the 3 ceramics with a cement space of 50 μm. A single pair of cylindrical and conical drills was used to mill all crowns of each system. After milling, the sprue was sectioned with a diamond disk, and adjustments were made with silicone stones, diamond stones, and rubber points (Zirconia and Lithium Disilicate Polishing Kit; Besser Dental). The LD and ZLS crowns were crystallized and Trans YZ sintered by following each manufacturer’s guidelines.
Before luting, the preparations and monolithic crowns were cleaned with distilled water in an ultrasonic bath (1440 DA; Odontobras) for 5 minutes and gently air-dried. The surfaces of the preparations were etched with 5% hydrofluoric acid (IPS Ceramic Etching Gel; Ivoclar Vivadent AG) for 60 seconds, followed by rinsing in water for 30 seconds and air-drying. Primers A + B (proportion 1:1) of the luting system (Multilink Automix; Ivoclar Vivadent AG) were mixed and applied to the preparations as per the manufacturer's recommendations.
The LD and ZLS crowns were etched by using 5% hydrofluoric acid (IPS Ceramic Etching Gel; Ivoclar Vivadent AG) for 20 seconds, followed by thorough rinsing in water for 30 seconds and air-drying. The silane bonding agent (Monobond Plus; Ivoclar Vivadent AG) was actively applied for 60 seconds, followed by gentle air-drying. Meanwhile, the intaglio surface of the Trans Yz crowns was submitted to airborne-particle abrasion with 45-μm aluminum oxide (Al 2 O 3 ) for 15 seconds at a 10-mm distance and 90-degree incidence angle under oscillatory motion to cover all surfaces, followed by gentle air-drying, active application of the silane bonding agent (Monobond Plus; Ivoclar Vivadent AG) for 60 seconds, and gentle air-drying again.
The crowns were cemented to their respective preparations with a dual-polymerizing resin-based luting agent (Multilink Automix; Ivoclar Vivadent AG). The paste base and catalyst were mixed in a 1:1 ratio, the crowns were filled and positioned over their preparations, and a load of 7.5 N was applied for 10 minutes. The cement excess was removed with a microbrush and then light polymerized (1200 mW/cm 2 – Radii-Cal; SDI) for 20 seconds in each of the 4 directions (0, 90, 180, and 270 degrees) and at the occlusal surface. The specimens were stored in distilled water in an oven (#502; FANEM) at 37 °C for 72 hours before fatigue testing ( Fig. 1 ). ,
Fatigue testing was executed with a step-stress methodology. An electric testing machine (ElectroPuls E3000; Instron) was used in which each crown (n=15) was positioned centrally over a flat metallic base inside a cylinder matrix to ensure standardized positioning. The assemblies were submerged in water, and a 40-mm-diameter stainless-steel sphere applied the load at the center of the crowns. Adhesive tape (110 μm) was positioned between the load applicator and the crown to enhance the contact between them and avoid contact damage (Hertzian cone cracks). Then, cyclical loads were applied at 20 Hz starting from 200 N for 5000 cycles to adjust the testing assembly, followed by sequential increments with a step size of 100 N at each 10 000 cycles until failure. The specimens were detached from the base and submitted to light transillumination to search for cracks after each testing step. If cracks were detected, the specimens were considered as failed, and data regarding the fatigue failure load and number of cycles for failure were recorded. However, if cracks were not detected, the specimens were repositioned and the testing continued.
After failure, each crown was submitted to fractography analysis. First, an optic stereomicroscope (Discovery V20; Carl Zeiss AG) was used to inspect the region of failure and the direction of crack propagation. Then, representative specimens were selected (n=3) and sectioned perpendicularly to the crack by using a diamond blade in a precision cutting machine under constant water cooling (Isomet 1000; Buehler) and submitted to scanning electron microscopy (SEM) analysis (Vega3; Tescan) at ×150 and ×500 magnification. In addition, the cement thickness at the occlusal intaglio surface was measured at 3 different regions in 5 additional luted crowns of each group by using a similar methodology.
A complementary analysis in an atomic force microscope (AFM) was performed for 3 additional specimens of each group that were not luted to the dentin analog to assess the topographic features related to the microstructure and processing of each ceramic material. For that, the crowns were sectioned with a precision cutting machine (Isomet 1000; Buehler) to expose their intaglio surface at the occlusal part. AFM (Park NX10; Park Systems) enabled a nanometric scale evaluation executed through the noncontact mode and by using a reflective aluminum backside-coated highly doped monolithic silicon probe (Nanosensors PPP-NCHR; Park Systems) at an area of 5×5 μm. Data analysis was performed by using a software program (Park XEI v4.3.4Build22.RTM1; Park Systems) considering 2 outcomes: roughness (Ra parameter) and the fractal dimension, both obtained at 5 different regions of each crown, followed by the assessment of the average of all these measurements. For fractal dimension estimation, the box counting method was used.
With regard to data analysis, fatigue failure load and cycles for failure were submitted to survival analysis by means of the Kaplan-Meier and Mantel-Cox post hoc tests with a statistical software program (IBM SPSS Statistics, v21; IBM Corp) (α=.05). Survival rates were calculated for both parameters at the different testing steps. A Weibull analysis of such data was run (SuperSMITH; Fulton Findings) to assess the structural reliability of each evaluated ceramic by obtaining the Weibull moduli and its respective 95% confidence interval for both outcomes. Statistical differences for the Weibull moduli were obtained by using the maximum likelihood approach where the overlap of confidence interval indicates statistical similarities and where its absence points to statistical differences.
In addition, data of complementary evaluations (roughness, cement thickness, and fractal dimension) were first submitted to a descriptive analysis to obtain mean and standard deviation values. After ensuring homoscedastic and parametric distribution, 1-way ANOVA and Tukey post hoc tests were run for each outcome. Topographic and fractography features depicted in AFM and SEM micrographs, respectively, were also qualitatively evaluated.
The Trans YZ material showed better fatigue performance ( P <.001), followed by ZLS and LD ( Table 3 ; Table 4 ). However, these materials had statistically similar Weibull moduli ( Table 3 ), which indicated similar data spread (statistical structural reliability). Fractography analysis corroborated all failures as radial cracks ( Fig. 2 ). With regard to the thickness of the occlusal resin-based cement, SEM micrographs shown similarity for the 3 materials considered ( P >.05; F=35.344) ( Table 3 and Fig. 3 ).
|Groups||Fatigue Results||Complementary Analysis c|
|FFL (N) a||Weibull Moduli b||CFF a||Weibull Moduli b||Fractal Dimension||Ra (nm)||Cement Thickness (μm)|
|LD||987 (915-1058) C||9.04 (6.00-13.64) A||73 667 (66 544-80 790) C||6.72 (4.44-10.17) A||2.242 (0.077) A||226 (103) A||123 (33) A|
|ZLS||1187 (1063-1310) B||6.44 (4.20-9.89) A||92 613 (80 497-104 729) B||5.01 (3.26-7.70) A||2.238 (0.054) A||169 (33) B||115 (39) A|
|Trans YZ||1740 (1562-1918) A||5.21 (3.59-7.58) A||149 000 (131 180-166 821) A||4.51 (3.10-6.57) A||2.147 (0.210) B||93 (31) C||105 (40) A|
|400/15 000||500/25 000||600/35 000||700/45 000||800/55 000||900/65 000||1000/75 000||1100/85 000||1200/95 000||1300/105 000||1400/115 000||1500/125 000||1600/135 000||1700/145 000||1800/155 000||1900/165 000||2000/175 000||2100/185 000||2200/195 000||2300/205 000||2400/215 000|
|LD||1||1||1||0.93 (0.1)||0.80 (0.1)||0,67 (0.1)||0.40 (0.1)||0.07 (0.1)||0.0||-||-||-||-||-||-||-||-||-||-||-||-|
|ZLS||1||1||1||0.93 (0.1)||0.87 (0.1)||0.80 (0.1)||0.73 (0.1)||0.60 (0.1)||0.53 (0.1)||0.33 (0.1)||0.07 (0.1)||0.0||-||-||-||-||-||-||-||-||-|
|Trans YZ||1||1||1||1||1||1||1||1||1||1||0.73 (0.1)||0.67 (0.1)||0.40 (0.1)||0.40 (0.1)||0.27 (0.1)||0.27 (0.1)||0.27 (0.1)||0.13 (0.1)||0.13 (0.1)||0.13 (0.1)||0.0|
The remaining complementary analysis corroborates different topographic features for the tested materials, which were caused by their different microstructures interacting with the specific processing related to each material ( Fig. 4 ). At a nanometric scale, AFM micrographs depicted LD as having the roughest Ra surface ( P <.001; F=5.677), followed by ZLS and then Trans Yz (the smoothest surface) ( Table 3 ). Fractal dimension data corroborated that the surface complexity was also higher in LD ( P <.001; F=9.918), which was statistically similar to ZLS; therefore, the lowest topography complexity was observed for Trans YZ ( Table 3 ).